January 2025 Summaries

Apache Superset, a prominent tool for interactive analytics in business intelligence, faces challenges in maintaining performance as datasets grow in complexity. Traditional pre-computation pipelines, often used to accelerate OLAP queries, introduce complexity and inefficiencies, prompting a need for a more streamlined approach. StarRocks, a high-performance MPP OLAP database, offers an innovative solution through its materialized views, which dynamically optimize queries without requiring changes to existing BI workflows or SQL queries. By leveraging StarRocks' cost-based optimizer and seamless query rewrite capabilities, users can achieve sub-second query performance on BI dashboards while reducing engineering overhead and accelerating development cycles. This approach allows dashboards to be built directly on raw data, enabling on-demand query optimization and integration into existing workflows without altering architecture, thereby enhancing the overall efficiency and responsiveness of Apache Superset environments.

Preset has introduced a new Interactive Table feature within its enterprise offering of Apache Superset, enhancing how users explore and interact with tabular data. Powered by AG Grid technology, this tool offers advanced filtering, flexible aggregations, seamless data transfer, and robust performance to handle large datasets efficiently. It allows easy column adjustments and supports existing functionalities like time shift comparison and conditional coloring, making it a suitable upgrade from existing table visualizations. This development reflects Preset's dedication to providing enterprise-grade data exploration capabilities, while maintaining its user-friendly approach, and signifies its ongoing contribution to the open-source community, with hints of future enhancements like Pivot Tables on the horizon.

Drawing from extensive experience as an on-call and DevOps engineer, the text explores the complexities and considerations of running Apache Superset at scale, particularly highlighting the challenges faced when transitioning from internal to public-facing services. It emphasizes the importance of understanding the expanded threat surface and security posture changes that accompany this transition. The author shares insights into the operational demands of deploying Superset, such as managing Python web applications, React frontends, databases, and task processing systems, while stressing the necessity of expertise in troubleshooting and system optimization for efficiency. The text underscores the benefits of specialized knowledge and infrastructure optimization, which can lead to significant cost savings and efficiency gains, as demonstrated by Preset's approach to running Superset as a core focus. Ultimately, it advises organizations to carefully consider their capabilities and resources when deciding whether to manage Superset internally or opt for a managed solution, emphasizing that the priority should be on leveraging data for better decision-making rather than on managing technical infrastructure.

Preset's modern BI platform is now available as a managed service on Google Cloud Platform (GCP), offering organizations the flexibility to choose their cloud provider while maintaining high security and compliance standards. By using Preset Managed Private Cloud, users can operate multiple isolated Apache Superset workspaces within their own cloud environment, minimizing operational overhead as Preset handles platform management. The service includes automated deployments, upgrades, security patches, elastic scalability, and comprehensive management features such as role-based access controls and data encryption. While currently available on AWS and GCP, Preset is exploring potential support for Microsoft Azure, inviting interested parties to provide feedback to prioritize this development.